apache/druid · error · DruidException

'druid.coordinator.kill.period

Error message

'druid.coordinator.kill.period'[%s] must be greater than or equal to 'druid.coordinator.period.indexingPeriod'[%s]

What it means

At coordinator startup, DruidCoordinatorConfig.validateKillConfigs() checks that the kill-unused-segments cleanup period (druid.coordinator.kill.period) is not shorter than the main coordinator indexing period (druid.coordinator.period.indexingPeriod). Killing runs inside the indexing-period duty cycle, so a smaller kill period is meaningless and rejected as invalid operator input.

Solutions

Set druid.coordinator.kill.period to a value >= druid.coordinator.period.indexingPeriod (e.g. PT30M or larger, commonly P1D)
Or lower druid.coordinator.period.indexingPeriod so it does not exceed the desired kill period
Restart the Coordinator after fixing the properties in runtime.properties

Example fix

// before (runtime.properties)
druid.coordinator.kill.period=PT5M
druid.coordinator.period.indexingPeriod=PT30M
// after
druid.coordinator.kill.period=PT30M
druid.coordinator.period.indexingPeriod=PT30M

Defensive patterns

Strategy: validation

Validate before calling

final Period killPeriod = Period.parse(props.getProperty("druid.coordinator.kill.period", "P1D"));
final Period indexingPeriod = Period.parse(props.getProperty("druid.coordinator.period.indexingPeriod", "PT30M"));
if (killPeriod.toStandardDuration().getMillis() < indexingPeriod.toStandardDuration().getMillis()) {
  throw new IllegalArgumentException("druid.coordinator.kill.period must be >= druid.coordinator.period.indexingPeriod");
}

Try / catch

try {
  coordinator.start();
} catch (WebApplicationException | IllegalArgumentException e) {
  if (String.valueOf(e.getMessage()).contains("druid.coordinator.kill.period")) {
    // correct runtime.properties and restart
  }
}

Prevention

Validate kill.period >= period.indexingPeriod in config review/CI before deploy
Use ISO-8601 durations (PT30M, P1D) consistently in runtime.properties
Remember kill runs inside the indexing-period duty cycle; never set it shorter
Document customized indexingPeriod values per cluster to avoid copy-paste regressions

When it happens

Trigger: Starting the Coordinator with druid.coordinator.kill.period set to a duration strictly less than druid.coordinator.period.indexingPeriod (default 30m / PT30M).

Common situations: Operators tuning kill frequency after upgrading (kill config moved under druid.coordinator.kill.* KillUnusedSegmentsConfig), copying configs between clusters where indexingPeriod was customized, or setting kill.period=PT5M while leaving the default indexingPeriod.

Understand the failure class

Background: "Invalid value" and "allowed values are" config errors: what your library rejected and how to fix it — this error's family across 41 libraries.

Related errors

AI-assisted analysis of apache/druid@9b90983fd2 (2026-09-07). Data as JSON: /api/errors/a4d21e4b57ed50ec. Report an issue: GitHub.

Appendix: source

Thrown at server/src/main/java/org/apache/druid/server/coordinator/config/DruidCoordinatorConfig.java:104

  public HttpLoadQueuePeonConfig getHttpLoadQueuePeonConfig()
  {
    return httpLoadQueuePeonConfig;
  }

  private void validateKillConfigs()
  {
    validateKillConfig(killConfigs.auditLogs(), "audit");
    validateKillConfig(killConfigs.compactionConfigs(), "compaction");
    validateKillConfig(killConfigs.datasources(), "datasource");
    validateKillConfig(killConfigs.rules(), "rule");
    validateKillConfig(killConfigs.supervisors(), "supervisor");
    validateKillConfig(killConfigs.segmentSchemas(), "segmentSchema");

    // Validate config for killing unused segments
    final KillUnusedSegmentsConfig killUnusedConfig
        = killConfigs.unusedSegments(getCoordinatorIndexingPeriod());
    if (killUnusedConfig.getCleanupPeriod().getMillis() < getCoordinatorIndexingPeriod().getMillis()) {
      throw newInvalidInputExceptionForOperator(
          "'druid.coordinator.kill.period'[%s] must be greater than or equal to"
          + " 'druid.coordinator.period.indexingPeriod'[%s]",
          killUnusedConfig.getCleanupPeriod(), getCoordinatorIndexingPeriod()
      );
    }
    if (killUnusedConfig.getMaxSegments() < 0) {
      throw newInvalidInputExceptionForOperator(
          "'druid.coordinator.kill.maxSegments'[%s] must be a positive integer.",
          killUnusedConfig.getMaxSegments()
      );
    }
  }

  private void validateKillConfig(MetadataCleanupConfig config, String propertyPrefix)
  {
    if (!config.isCleanupEnabled()) {
      // Do not perform validation if cleanup is disabled
      return;

View on GitHub (pinned to 9b90983fd2)